Grinding wheel wear monitoring based on wavelet analysis and support vector machine

A novel grinding wheel wear monitoring system based on discrete wavelet decomposition and support vector machine is proposed. The grinding signals are collected by an acoustic emission (AE) sensor. A preprocessing method is presented to identify the grinding period signals from raw AE signals. Root mean square and variance of each decomposition level are designated as the feature vector using discrete wavelet decomposition. Various grinding experiments were performed on a surface grinder to validate the proposed classification system. The results indicate that the proposed monitoring system could achieve a classification accuracy of 99.39% with a cut depth of 10?μm, and 100% with a cut depth of 20?μm. Finally, several factors that may affect the classification results were discussed as well.     

砂轮磨钝监测及修整

利用声发射(AE）信号归原处理法对砂轮磨钝程度进行监测，在此基础上开发了砂轮磨钝监测及自动修整系统。该系统可以对小批量、多品种工件磨削过程中砂轮钝化进行有效的监测，采用的CNC砂轮自动修整器能够解决多种类型成型砂轮的修整，具有一定的通用性和实用性，能实现磨削砂轮监测、修整过程的自动化和智能化。     

磨削砂轮状态在线监测方法研究

提出了用声发射（AE）信号在线监测砂轮状态的方法.利用该方法可以监测工件材料、加工要求和磨削参数经常变化环境下的砂轮钝化程度和破碎情况;并采用神经网络建立了传感器信号与砂轮状态之间的非线性关系.     

Modeling of grinding wheel topography based on a joint method of 3D microscopic observation and embedded grindable thermocouple technique

The grinding wheel generally has a complicated topography for the irregularity of abrasive grits, which always has an important influence on the final quality of the grinding workpiece. In this paper, a joint method of microscopic observation and grindable thermocouple technique was adopted to model the wheel topography. The grinding wheel topography was first modeled through microscopic observation by an in-position 3D microscope KH-7700 installed on the grinding machine. Based on the measurement of grit sizes, shapes, and distributions through the 3D microscope, a wheel surface model was established and a static grit number model based on Rayleigh distribution was proposed. Moreover, a numerical model was given to validate the proposed Rayleigh distribution model of an active grit number. In order to investigate the real abrasive grit number in a grinding process, an embedded grindable thermocouple was used to detect the dynamic variation of temperature signals, which will reflect the variation of in-process wheel topography under different process parameters, machine status, and even the grit-workpiece interaction status. Through the experimental analysis, it can be concluded that the increase of depth of cut a_p could help to greatly increase the active grit number to the grinding process, while the increase of workpiece speed V_w and decrease of wheel speed V_s could lead to a subtle increase of the grit number. Moreover, the active grit number is about one fourth to one third of the static grits. The contact arc length between the wheel (CBN) and the workpiece (Ti-6Al-4V) was calculated by the contact time from the workpiece surface temperature data, and it was found that the actual contact arc length was about 1.5~2 times of the geometric size.     

砂轮精确修整时的声发射检测方法

砂轮修整状态严重地影响砂轮的磨削性能,这在精磨及成形磨削过程中显得尤为突出。本文提出了一种砂轮精确修整的新方法,利用声发射试验检测系统实时提取砂轮修整过程中声发射信号的特征参量,有效地实现了砂轮型面的精确修整。     

Process monitoring in precision cylindrical traverse grinding of slender bar using acoustic emission technology

The precision cylindrical traverse grinding process of slender bar is very complex for the strongly time dependent properties of the wheel. Therefore, it is very difficult for operators to properly judge the grinding state using naked eyes and ears. This calls for automatic monitoring technology that can monitor the process in precision cylindrical traverse grinding to guarantee machining quality and productivity as well as reduction in cost. This study developed an automatic monitoring system for precision cylindrical traverse grinding of slender bar using Acoustic emission (AE) technology. Grinding tests on molybdenum were conducted under traverse conditions in a conventional cylindrical grinder. It was found that larger radial material removal depth results in larger root mean square value of Acoustic emission signals (AE_RMS). Based on this, the AE_RMS was analyzed and used to determine the finishing of spark-out process and the pre-processing of tool alignment. The variation tendency of AE_RMS in one spark-out process was applied to determine when a wheel wears out and has to be dressed. The experimental results showed that the AE system was effective to monitor the pre-processing of tool alignment, spark-out and wheel wear in precision cylindrical traverse grinding of slender bar.

     

Stability analysis for SiC grinding based upon harmonic wavelet and Lipschitz exponent

Abstract

In order to effectively identify the stability degree of grinding and obtain better surface integrity of machined workpiece in grinding application, in this paper, a quantitative assessment approach based on harmonic wavelet and Lipschitz exponent is proposed, using silicon carbide as an example. To begin with, the Lipschitz exponent is defined and calculated by harmonic wavelet modulus maxima (HWMM) and linear regression algorithm, respectively. Then the Kurtosis based on Lipschitz exponent (KLE) is defined as a quantitative indicator. In grinding experiment, the acceleration signals and grinding force signals of the X/Y directions are selected, and experimental parameters including wheel speed, workpiece speed and grinding depth are employed to investigate the grinding stability in grinding stage. Experimental results show that the grinding stability is relatively worse at low wheel speed, low-high workpiece speed and large grinding depth in terms of KLE index, which may provide a new guidance to explore optimized grinding parameters and thus obtain good surface integrity in engineering applications.     

磨削砂轮状态在线监测方法研究

用声发射(AE)信号在线监测砂轮状态的方法,可以监测工件材料、加工要求和磨削参数经常变化环境下砂轮钝化程度和破碎;并采用神经网络建立传感器信号与砂轮状态之间的非线性关系.     

基于热像检测与分割技术的磨削砂轮堵塞检测研究

针对磨削加工中的砂轮堵塞检测的难题,提出了一种基于主动红外检测技术和热像图分割技术的砂轮堵塞检测方法。分别运用阈值分割、边缘检测和区域生长三种方法分割热像图,提取了砂轮堵塞量和状态的特征。实验结果表明：阈值分割法提取的砂轮堵塞磨屑的分布出现“偏聚”现象,堵塞特征受选取的阈值影响较大;边缘检测法克服了“偏聚”现象,但仍存在边缘不完整、定位不精确等问题;而基于梯度阈值过滤的区域生长分割法检测结果更客观和准确。把基于梯度阈值过滤的区域生长分割法的砂轮堵塞红外检测结果与显微照相法的砂轮堵塞检测结果进行了分析比较,结果表明基于梯度阈值过滤的区域生长分割的红外检测法可以实现对砂轮堵塞面积和状态的定量检测。     

磨削加工参数智能化在线调整方法研究

提出运用神经网络的非线性建模原理和自学习能力，建立影响磨削参数选择的各种因素与磨削参数之间的函数映射关系。据此可以实现初始磨削用量智能化选择目的。将磨削过程智能监测模块检测到的砂轮状态特征信号反馈到神经网络的输入端，根据砂轮钝化程度实现磨削参数的在线智能调整，以减少砂轮钝化对磨削质量的影响，确保加工质量的稳定性。     

DEVELOPMENT OF AN OPTICAL SCANNER TO STUDY WEAR ON THE WORKING SURFACE OF GRINDING WHEELS

The efficiency of the grinding operation is highly dependent on the grinding wheel surface topography. Several methods for the evaluation of the grinding wheel surface have been developed in the last few years. In some of these methods, the grinding wheel has to be removed from the machine for an evaluation using a microscope or a profilometer. Some other methods are able to measure the topography at the grinding machine, but they do not give detailed information about grain distribution and wear. This paper shows a new method to map the grinding wheel surface. The proposed system is based on an optical scanner capable of measuring the light beam reflected from the wear flat areas on the abrasive grains over the whole wheel peripheral surface. The Mapping System for Grinding Wheels (MSGW) is able to acquire data with the wheel running at the work speed (30 m/s)and the measurement carried out on the grinding machine without stopping. The system is applied on a surface grinding operation where an Alj0₃grinding wheel is used. The results show that the system can be used to map the wear flats at the grinding wheel surface and also to analyze the wear phenomena efficiently.     

Grinding with directionally aligned SiC whisker wheel-loading-free grinding

A new grinding wheel has been developed, wherein very fine SiC whiskers are aligned normally to the grinding surface. However, the grinding wheel exhibits a tendency toward loading. In this study, a method to improve grinding efficiency was proposed. The loading of the grinding wheel could be remarkably reduced by the addition of a small amount of very fine abrasive grains to the grinding fluid. The experimental results showed that alumina of 0.6-μm grain size is most suitable, and a concentration of 0.25% wt is sufficient to prevent the loading. The grinding efficiency and grinding ratio were remarkably improved by this method.     

Analytical modeling of grinding wheel loading phenomena

Wheel surface condition plays an important role in the grinding operation. Grinding wheel loading, meaning chip accumulation in the space between grains, leads to deteriorating wheel cutting ability and causes excessive force and temperature. This paper presents an analytical model of wheel loading phenomena as a function of cutting parameters, wheel structure, and material properties. The model is based on the adhesion of workpiece material to abrasive grain surface. It is validated by experimental results from grinding nickel-based superalloy with cubic boron nitride vitrified wheel. This model considers wheel specifications including abrasive grains size and the number of cutting edges. Cutting parameters and process temperature are the other determinant factors. On the basis of this model and empirical results, the effects of the various process parameters are presented.     

Reduction of Air-Grinding Time Using Collaboration of Dual Sensors

Air-grinding time is inevitable in the grinding process. Since the low feedrate during air grinding makes the process inefficient, air-grinding time should be reduced. This paper presents a technology for reducing air-grinding time in cylindrical plunge grinding operations. The main idea is to estimate a distance, NAP (nearest approaching point), between the workpiece surface and the position of the grinding wheel. After detecting the NAP using an ultrasonic sensor, a CNC controller adjusts the feedrate using the feedrate override function with the help of the AE signal that gives the CP (contact point) information. The experimental results show that an ultrasonic and an AE sensor are good enough for detecting the NAP and the CP, respectively, and the system reduces the conventional air-grinding time by two-thirds.     

Workpiece roundness profile in the frequency domain: an application in cylindrical plunge grinding

In grinding, most control strategies are based on the spindle power measurement, but recently, acoustic emission has been widely used for wheel wear and gap elimination. This paper explores a potential use of acoustic emission (AE) to detect workpiece lobes. This was achieved by sectioning and analysing the AE signal in the frequency domain. For the first time, the profile of the ground workpiece was predicted mathematically using key frequencies extracted from the AE signals. The results were validated against actual workpiece profile measurements. The relative shift of the wave formed on the surface of the part was expressed using the wheel-workpiece frequency ratio. A comparative study showed that the workpiece roundness profile could be monitored in the frequency domain using the AE signal during grinding.     

Experimental study of burn classification and prediction using indirect method in surface grinding of AISI 1045 steel

Grinding burn is a discoloration phenomenon according to the thickness of oxide layer on the ground surface. This study tries to establish an automatic grinding burn detection system with robust burn features that are caused by burn and not by the design parameters. To address this issue, a method based on acoustic emission sensor, accelerator, electric current transducers, and voltage transducers was proposed in an attempt to extract burn signatures. A trial-and-error experimental procedure was presented to find out burn threshold. Vitrified aluminum oxide grinding wheel and AISI 1045 steel workpiece were used in the grinding test, as they were the most commonly used wheel–workpiece combinations in conventional grinding process. With the help of fast Fourier transform and discrete wavelet transform, the spectral centroid of AE signal, the maximum value of power signal, and the RMS of the AE wavelet decomposition transform from wavelet decomposition levels d1 to d5 were extracted as burn features. The spectral centroid of AE signal was believed not to be affected by grinding parameters. A classification and prediction system based on support vector machine was established in order to identify grinding burn automatically. Results indicate that the classification system performs quite well on grinding burn classification and prediction.     

Image processing of grinding wheel surface

Identifying the situation of grinding wheel wear and loading is a very important issue for high efficiency grinding operations. This paper presents a new method that detects and identifies the chip loading and cutting edge wear of a grinding wheel using the image processing toolbox of MATLAB. The different optical characters of the metal chips and the abrasive grains are analysed. The Sobel operator is adopted to make edge detection. A sensitivity threshold based on the global condition is used to decrease the noise. Image dilation and erosion processes are used to ensure the edge of each loaded chip is covered by a continuous section. The ratios of chips are calculated and displayed to monitor the wheel surface working status.     

Deformation analysis of micro/nano indentation and diamond grinding on optical glasses

The previous research of precision grinding optical glasses with electrolytic in process dressing (ELID) technology mainly concentrated on the action of ELID and machining parameters when grinding,whic...     

Image processing of the grinding wheel surface

Identifying the situation of grinding wheel wear and loading is a very important issue for high-efficiency grinding operations. This paper presents a new method that detects and identifies the chip loading and cutting edge wear of a grinding wheel using the image processing toolbox of the MATLAB package. The different optical characters of the metal chips and the abrasive grains are analysed. The Sobel operator is adopted to make the edge detections. A sensitivity threshold based on the global condition is used to lower the noise. Image dilation and erosion processes are used to ensure that the edge of each loaded chip is covered by a continuous section. The ratios of chips are calculated and displayed to monitor the wheel surface working status.     

磨削技术的新进展──硬脆材料光滑表面的超精磨削

阐述利用超细粒度金刚石砂轮进行硬脆材料光滑表面起精磨创的现状，提出获得硬脆材料光滑磨削表面的主要技术措施。